KeyArgo
/
youtube-transcript-api
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

move _FORMATTING_TAGS inside _TranscriptParser

This commit is contained in:
E. Seiver 2023-04-13 12:27:34 -07:00
parent eda8ddb38f
commit 79fd63d585
1 changed files with 13 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
youtube_transcript_api/_transcripts.py
View File

@ -27,19 +27,6 @@ from ._errors import (
) )
from ._settings import WATCH_URL from ._settings import WATCH_URL
_FORMATTING_TAGS = [
'strong', # important
'em', # emphasized
'b', # bold
'i', # italic
'mark', # marked
'small', # smaller
'del', # deleted
'ins', # inserted
'sub', # subscript
'sup', # superscript
]
def _raise_http_errors(response, video_id): def _raise_http_errors(response, video_id):
try: try:
@ -341,11 +328,23 @@ class Transcript(object):
class _TranscriptParser(object): class _TranscriptParser(object):
def __init__(self, preserve_formatting=False): def __init__(self, preserve_formatting=False):
self.preserve_formatting = preserve_formatting self.preserve_formatting = preserve_formatting
self._FORMATTING_TAGS = [
'strong', # important
'em', # emphasized
'b', # bold
'i', # italic
'mark', # marked
'small', # smaller
'del', # deleted
'ins', # inserted
'sub', # subscript
'sup', # superscript
]
self._html_regex = self.get_html_regex() self._html_regex = self.get_html_regex()
def get_html_regex(self): def get_html_regex(self):
if self.preserve_formatting: if self.preserve_formatting:
formats_regex = '|'.join(_FORMATTING_TAGS) formats_regex = '|'.join(self._FORMATTING_TAGS)
formats_regex = r'<\/?(?!\/?(' + formats_regex + r')\b).*?\b>' formats_regex = r'<\/?(?!\/?(' + formats_regex + r')\b).*?\b>'
html_regex = re.compile(formats_regex, re.IGNORECASE) html_regex = re.compile(formats_regex, re.IGNORECASE)
else: else: